38 EFFECTS OF IONIZING RADIATION peared to show a more consistent pattern than did the leukocyte counts. Platelet counts on one individual, considered as a typical response in a non-fatal Japanese (13) indicated an apparent low approximately on day 30. This time trend agrees with that seen in the Marshallese and Americans exposed to fallout radiation. It is worthy of note that the period of peak incidence of purpura in the Japanese victims occurred between the 25th and 30th day, which corresponds to the time of maximum platelet depression in the exposed Marshallese. 4.43 Comparison With Data From Laboratory Accidents Although in the Los Alamos (18) and Argonne accidents (19) the type of radiation and the conditions of exposure were markedly different from either the Japanese or the Group I situations, a large component of penetrating gamma and neutron radiation was received and thus attempts at comparison may be of value. Somefindings in the hematological responses are pointed out: a) a uniform early rise in white and neutrophile counts over the first few days, similar to that seen early in the American group was observed uniformly.* b) of three high-exposure but non-lethal cases, the total white and leukocyte counts continued to show some degree of depression into the 7th week or beyond. c) the lymphocyte counts in individuals exposed to as little as 50 rem showed an initial marked depression. In most cases the lymphocyte counts remained at low levels throughout the period of observation. d) plate- The .Argonne Laboratory accident (19) involved four individuals who were estimated to have received 136, 127, 60 and 9 rep, respec- tively. The tindings in the two highest exposed individuals in general were consistent with those in the present study. An initial neutrophilic leukocytosis was followed by fluctuations in total count, with low values continuing into the 7th week. Recovery was not complete by the 20th week. The lymphocyte depression was rapid and marked, recovery was not evident by the 20th week. Minimumvalues for the platelet counts were obtained between the 25th and Bist day. Comparison with Animal Data. The time trends and severity of peripheral blood count change following total body radiation in animals has been examined critically recently (15), and the following general conclusions are presented. a) .An initial rise in total white count (re- flected in the neutrophile count) may occur. Thereafter the magnitude of depression of the total white and neutrophile counts, and within limits their duration are a function of radiation dose. .4 secondary or abortive rise in the total white count (reflected in the neutrophile or lymphocyte count) may occur, followed by a second decrense. There is little species difference in the rate of depression of the total white or neutrophile count at comparable doses; however, the rate of recovery and time for complete recovery is quite different in various species. Small animals (mouse, rat, hamster) showrelatively complete recovery to control levels, even at doses in the lethal range, by the end of the oth week or earlier. Data on dows are inade- let counts were done by a different method, and absolute counts are therefore not compurable. However, of the three high dose survivors, times of maximumdepression were not inconsistent with the value of 30 days obtained in the present studies. In higher dose non-survivors, however, the platelet counts had reached minimum values as early us the sth day. quate to indicate when recovery is complete; however, return to control levels at high dose levels has not occurred by the 5th week. Swine require 9 to 15 or more weeks for complete recovery. b) The response of lymphocytes is essentially identical in all animal species. Depression carn be detected within a few hours, and recovery *No counts were taken on Groups [ and © during the first T2 hours. recovery. from the minimum values (achieved in 36 to 48 hours) requires longer chan does neutrophile Lymphocytes fall to very low levels